Combined treatment of submacular hemorrhage with low-dose subretinal recombinant tissue plasminogen activator and intravitreal conbercept

Abstract

Background

Pars Plana Vitrectomy (PPV) combined with subretinal injection of low-dose recombinant tissue plasminogen activator (rt-PA) and intravitreal injection of Conbercept as a novel therapy for submacular hemorrhage (SMH) requires evaluation.

Methods

In a retrospective interventional clinical study, 14 eyes of 14 patients with SMH underwent PPV along with rt-PA (subretinal) and Conbercept (intravitreal) injections. The main outcomes included best-corrected visual acuities (BCVAs), degrees of blood displacement, and adverse events. All patients completed at least 6-month follow-up visits.

Results

Mean BCVAs significantly improved at 7 days (22.29 ± 15.35), 1 month (30.71 ± 16.42), 3 months (38.29 ± 13.72), 4 months (38.86 ± 14.15), and 6 months (41.21 ± 14.91) post-treatment compared to baseline (16.36 ± 13.97) (F = 12.89, P = 0.004). The peak improvement in BCVAs occurred at 6 months postoperatively. The procedure effectively eliminated subfoveal hemorrhages in all eyes, with clots removal and absorption occurring within one month and complete regression by 3-month follow-up visits. Postoperatively, two cases of AMD resulted in discoid scars on the fundus. No instances of rt-PA-related retinal toxicity were observed during the follow-up period.

Conclusion

The combined approach of PPV with low-dose rt-PA and anti-VEGF shows promise in enhancing both vision and anatomical structure in SMH therapy. Individualized treatment plans tailored to the primary disease should be developed to optimize visual prognoses.

Trial registration

Retrospectively registered No.ChiCTR2100053034. Registration date: 10/11/2021.

Background

Submacular hemorrhage (SMH) is a severe, vision-threatening condition with multiple potential causes [1,2,3,4,5]. As a leading cause of blindness worldwide among the working population, SMH affects approximately 16.5% of individuals [6, 7]. Factors like blood hemosiderin toxicity to photoreceptors, subretinal blood impeding the Retinal Pigment Epithelium (RPE), compromised nutrient diffusion in the choroid, and the underlying pathology contributing to bleeding all negatively impact visual prognosis [8].

Bleeding in the fovea precipitates sudden and progressive vision loss, resulting in irreversible visual impairment [2]. Therefore, early intervention in SMH is crucial to mitigate retinal damage [9, 10]. Although retrospective reports suggest functional benefits from treatments like photodynamic therapy (PDT), anti-vascular endothelial growth factor (anti-VEGF) drugs, and pars plana vitrectomy (PPV) with thrombectomy when administered promptly after acute SMH [2, 11], the consensus on optimal therapy remains unclear [12,13,14].

Anti-VEGF agents, as the primary treatment modality, effectively prevent rebleeding and neovascularization but are ineffective in removing hemorrhage due to short half-life periods, necessitating repeated injections that increase healthcare burdens and the risk of endophthalmitis [2, 6, 7].

Recombinant tissue plasminogen activator (rt-PA), proposed by Heriot as an enzyme for SMH treatment, degrades fibrin and certain coagulation factors to dissolve hemorrhages [9, 15, 16]. Early retrospective case series have confirmed promising results, particularly with subretinal rt-PA injection [8,9,10, 16, 17]. Haupert et al. optimized the method by adding PPV to mitigate potential photoreceptor degeneration and prevent severe complications like retinal detachment [1, 17, 18]. However, controversies regarding PPV-related postoperative complications and the retinal toxicity of rt-PA persist [19, 20].

Since anti-VEGFs and rt-PA operate via different mechanisms, their simultaneous application targets various pathophysiological mechanisms of SMH, potentially yielding better visual and functional outcomes [2, 12, 21].

As few studies have explored the combination therapy of anti-VEGFs and rt-PA, this research aims to evaluate the effectiveness and safety of PPV with low-dose subretinal rt-PA and intravitreal anti-VEGF injection for SMH treatment [21,22,23]. Furthermore, this study investigates how personalized treatment plans affect the visual acuity of patients during extended follow-up appointments.

Methods

Study design and patients

The retrospective interventional single-arm study received approval from the Ethics Committee of Tongji Hospital and adhered to the principles of the Declaration of Helsinki. All data were completely anonymized before access. Medical records of 14 patients diagnosed with SMH who underwent vitrectomy combined with subretinal injection of rt-PA and intravitreal injection of Conbercept between July 2021 and September 2023 were reviewed. SMH was defined as a noticeable subfoveal retinal elevation observed during biomicroscopic examination [2, 23]. Eligible patients were aged over 18 years, had treatment-naïve SMH (2–5 disc diameters), and central subfield hemorrhage thickness (CST) > 250 μm measured by optical coherence tomography (OCT). Patients undergone previous ocular treatments, including intravitreal injections or retinal laser photocoagulation within the last 3 months, or had a history of vitrectomy or PDT, were excluded.

After obtaining written informed consent, patients underwent a comprehensive assessment, including evaluation of disease duration, and history of ocular and systemic diseases such as diabetes mellitus and systemic hypertension. Ophthalmologic examinations included Spectra-Domain OCT imaging (Spectralis; Heidelberg Engineering GmbH, Heidelberg, Germany), ultra-wide-field color fundus photography (Daytona; Optos plc, Scotland, United Kingdom), Early Treatment Diabetic Retinopathy Study (EDTRS) best-corrected visual acuity (BCVA) test, intraocular pressure (IOP) assessment, and detailed fundus examinations (NT-510; NIDEK, Honshu, Japan). Follow-up visits continued for at least 6 months and were conducted by a masked expert specialist.

For Spectra-Domain OCT (SD-OCT), 5 horizontal lines followed by 5 vertical lines with a 6-mm scan length and 0.5-mm interval were performed [24]. The protocol of central retinal thickness (CRT) map analysis within ETDRS subfields was applied. CST was defined as the average CRT with a 1 mm diameter [25]. Blood displacement was defined as clot removal within 1 DD of the fovea on SD-OCT [26] .

The primary outcome included the average change in amplitude of BCVAs at the 6th month compared to the baseline. Secondary outcomes included the displacement of submacular clot on OCT and the incidence rate of ocular complications during follow-up.

Treatment protocol

The surgery proceeded as follows. An expert surgeon performed PPV using the 23-gauge (G) microincision technique to induce a posterior vitreous detachment. A 38G needle was used for the subretinal injection of rt-PA. The ultra-fine needle minimally penetrated the retina and delivered a precise amount, not exceeding 25 µg/0.05mL, to adequately fill the bleeding area. Sterile air was injected into fill the vitreous cavity for retinal reattachment. Conbercept (Chengdu Kanghong Biological Co., Ltd., Chengdu, China) was finally injected intravitreally as a single 0.5 mg dose.

Notably, the internal limiting membrane peeling (ILMP) was innovatively performed before injections to release vitreomacular adhesions and reduce tractions at the vitreoretinal interface. The approach aims to decrease the recurrence of macular disease and vitreomacular traction syndrome postoperatively. To minimize iatrogenic mechanical damage to the retina, we have innovated injection procedures by administering rt-PA immediately when the cannula punctures the highest point of the hemorrhage bulge.

Postoperative personalized treatment

Monthly follow-up visits were conducted. Re-treatments of surgical eyes were evaluated using OCT imaging and fundus photography. Conbercept injections were administered pro re nata (PRN) if subretinal fluid or recurrent SMH occurred with increasing CST > 50 μm or loss of EDTRS > 5 letters.

Data analysis

BCVAs were recorded as EDTRS letters for statistical analysis. Normal distribution of data was assessed using the Shapiro-Wilk test and Q-Q plot. To compare preoperative and postoperative BCVAs, the Kruskal-Wallis test, Mann-Whitney test and repeated measures ANOVA analysis of variance were used to assess the changes in VA and IOP during the follow-up period. Statistically significant was defined as a p-value less than 0.05. Data analyses were performed using SPSS software version 25.0 (SPSS, Inc., Chicago, Illinois, USA), and results are presented as mean ± standard deviation (SD).

Results

Patient demographic information

Demographic information of all patients is summarized in Table 1. The most prevalent diagnosis was PCV (42.9%, 6/14), followed by AMD (28.6%, 4/14). The mean age (± SD) of the 6 male and 8 female patients was 59.64 ± 14.80 years (range, 29–83 years). The duration from onset to operation was 18.07 ± 7.89 days (range, 6–32 days), with a mean follow-up time of 11.36 ± 5.93 months (range, 6–22 months). No postoperative complications were reported during follow-up visits.

Table 1 Characteristics of patients with SMH

Postoperative visual results

Mean BCVAs at 7 days (22.29 ± 15.35), 1 month (30.71 ± 16.42), 3 months (38.29 ± 13.72), 4 months (38.86 ± 14.15), and 6 months (41.21 ± 14.91) after treatment showed significant improvement compared to baseline BCVAs (16.36 ± 13.97) (F = 12.89, P = 0.004) (Fig. 1). The average improvement in final vision was (24.85 ± 17.06). 6 out of 14 eyes (42.9%) achieved a final vision of more than 30 letters, while 1 out of 14 eyes (7.1%) experienced lost vision by the final visit. No significant correlations were found between final visual acuity and age, diagnosed disease, disease duration, follow-up time, or preoperative visual acuity.

Fig. 1

The BCVA changes across follow-up visits. Scattering represented mean BCVAs. Postoperative vision gradually improved and stabilized over time with follow-up appointments

Postoperative condition of affected eyes

The anatomical structure changes are illustrated in Figs. 2, 3, 4 and 5. Blood clots were almost completely removed from the fovea and absorbed within one month after surgery. Fundoscopic examination revealed complete clearance of submacular hemorrhage at the final visit.

Fig. 2

Multimode imaging of SMH displacement in patient 1, a 33-year-old man with PM. A1-A2. (A1) ultra-wide color fundus at the initial visit showed SMH. (A2) SD-OCT showed raised hemorrhage under the fovea. B1-B2. (B1) Fundus photos taken 1 month after surgery show clearance of SMH. (B2) SD-OCT 1 month after surgery shows the disappearance of SMH and disrupted the Ellipsoid Zone

Fig. 3

Multimode imaging of SMH displacement in patient 5, a 59-year-old woman with PCV. A1-A2. (A1) ultra-wide color fundus at baseline showed SMH. (A2) SD-OCT showed dense hemorrhage. B1-B2. (B1) ultra-wide color fundus at 1 week postoperatively showed clearance of SMH. (B2) SD-OCT 1 month after surgery shows removal of SMH and disrupted Ellipsoid Zone

Fig. 4

Multimode imaging of SMH displacement in patient 6, a 63-year-old man with RAM. A1-A2. SD-OCT showed dense hemorrhage under the fovea. B1-B2. SD-OCT 1 month after surgery showed regression of SMH and changes in the Ellipsoid Zone

Fig. 5

Multimode imaging of SMH displacement in patient 7, a 64-year-old man with PCV. A1-A2. SD-OCT showed dense hemorrhage under the fovea. B1-B2. SD-OCT 1 month after surgery showed regression of SMH and pigment epithelium detachment

Hemorrhage had completely regressed in all patients by the 3-month follow-up. Three patients with PCV and one with PM exhibited retinal neovascularization at the 1-month visit and received Conbercept injections as therapy. Two patients with AMD demonstrated RPE degeneration one month postoperatively. Patients with AMD and macular edema received Conbercept injections three months postoperatively. All other patients received treatment as needed, with no recurrence of SMH or complications such as macular holes and retinal detachments occurring during the visits.

Discussion

Submacular hemorrhage (SMH) typically arises from choroidal or retinal circulation abnormalities in the macular, leading to blood accumulation beneath or between the retina and RPE [8, 18, 27]. Irreversible retinal damage and degeneration of photoreceptor function occur after hemorrhage in animal models [2, 19]. Various mechanisms contribute to visual function impairment due to SMH, including hindered diffusion of nutrients and oxygen, mechanical traction damage to photoreceptors, retinal toxicity from iron ions, and progressive damage from subretinal fibrosis and disciform scars [2, 12, 28]. Glatt et al. proposed that tissue lesions occur within 24 h and result in complete photoreceptor damage at 1 week in SMH [19]. Nonetheless, a retrospective study evaluating natural history of SMH concluded the efficacy of surgery to prevent severe vision loss [29]. Bennett et al. confirmed this and noted that patients with thick SMH tend to have poorer visual prognoses, and dense macular hemorrhage can obscure underlying lesions, impeding effective treatment [30]. Therefore, early clearance of subretinal or sub-RPE blood is recommended [9, 10].

Early treatment modalities involve intravitreal medications and vitrectomy, often combined with intravitreal or subretinal injections [2, 14, 31]. Conbercept as one anti-VEGF has recently obtained approval from FDA in China and is safe and effective with lower costs [31]. While anti-VEGFs are the first-line treatment since their effectiveness in preventing neovascularization and leakage, they fail to remove SMH clots [6]. Trials specifically assessing anti-VEGF agents for SMH also lack. Although the safety of anti-VEGF monotherapy is proven, reports about recurrent SMH still exist [2]. Short half-life periods also necessitate repeated injections that increase burdens and risks of endophthalmitis [32, 33].

A revised therapy proposed by Haupert et al. involves vitrectomy with subretinal rt-PA injection and gas tamponade, aiming for more effective clot dissolution [17]. rt-PA promotes conversions of fibrinolytic enzymes that degrade fibrin to dissolve clots [2, 8, 34]. However, this approach may lead to postoperative complications such as re-hemorrhage and retinal breaks [8, 30, 35]. Single therapeutic pathways are insufficient for SMH treatment, necessitating combination therapy for improved vision and anatomical outcomes. Combined therapy of PPV, intravitreal anti-VEGFs and subretinal rt-PA appears to yield positive outcomes, though such combinations have not been extensively studied [22, 36].

Our study introduces combination therapy for SMH treatment, along with procedural enhancements to minimize iatrogenic damage. Additionally, we enhanced the subretinal injection procedure by using an ultrathin 38G needle and carefully advancing the syringe only when the cannula punctures the hemorrhage bulge, thus minimizing iatrogenic lesions. Surgical procedures were also innovated. ILMP was performed to release vitreomacular adhesions, reducing the risks of retinal detachment and vitreomacular traction syndrome. We utilized sterile air tamponade and a prone position postoperatively to minimize bleeding and displace liquefied clots away from the fovea. The approach is advantageous for the self-sealing of iatrogenic retinal tears. Anti-VEGF injections at the end of surgery were advocated to induce regression of potential neovascularization or polyps concealed by dense SMH [37]. We observed regression of subretinal blood in all eyes within one month postoperatively, with no vitreoretinal complications. In addition, the main functional structures recovered after therapy. All patients had disrupted areas of Ellipsoid Zone at baseline and one PCV patient was found partially Ellipsoid Zone repaired, less than results from Limon et al. [38]. Nevertheless, by the three-month follow-up, all patients demonstrated complete removal of hemorrhage and restoration of normal layered anatomical structures including PRE on SD-OCT, which suggested the positive effect on the recovery of main anatomical structures and prevention of further functional deterioration following bleeding [23]. Visual acuity also gradually improved over time, peaking at six months postoperatively. Our results suggest the relative safety and efficacy of combination therapy, indicating its potential as a treatment approach.

Controversies exist regarding the retinal toxicity of rt-PA. Animal studies have found toxicity when the dosage of intravitreal injection exceeds 50 µg, and retinal exudative detachment occurred with doses over 100 µg. The vitreous volume of these animals is significantly smaller, indicating a higher concentration of rt-PA in the vitreous compared to the equivalent dose in humans. In this study, patients underwent a lower rt-PA dosage (25 µg). with no observed retinal toxicity, consistent with reports from Hassan et al. and Tsai et al. [35, 39].

In our study, We observed that two AMD patients developed RPE degeneration following hemorrhage lysis, marked by macular disciform scars. The phenomenon could be attributed to the proliferation of retinal fibers and vessels in the primary disease, along with the prevalence of discoid scars [3]. Additionally, three patients with PCV and one with PM exhibited retinal neovascularization a month post-surgery. This occurrence might be linked to the likelihood of dense SMH concealing potential lesions and restricting the information gleaned from angiography [34, 40]. The remaining patients experience no postoperative complications during the follow-up period.

Postoperative visual prognosis is influenced by various factors including the underlying cause, macula condition pre-bleeding (present edema, neovascularization, or tears), and bleeding characteristics (time, size, extent) [3, 8, 30, 34]. The efficacy and prognosis can vary based on individual circumstances. AMD accounts for 80% of cases and tends to have a poorer prognosis due to abnormal proliferation and fibrosis of more subretinal blood vessels [11, 29], which our study corroborates. However, we found no significant correlation between final vision and factors such as age, diagnosed disease, disease duration, follow-up time, or preoperative vision. This contrasts with the findings findings of Lin et al. [3] and underscores the need for further studies with larger sample sizes.

For SMH patients, personalized treatments are implemented based on the progression of retinal leakage or neovascularization as observed through ocular examinations. In this study, PCV patients were evaluated and treated with Conbercept on an as-needed basis to eliminate potential polyps, decrease exudative fluid, and mitigate further harm from recurrent conditions [21, 31, 32]. The results showed that eyes receiving this treatment had improved visual outcomes compared to their initial vision. Therefore, the study suggests that personalized treatment plans should be continued to maintain postoperative vision based on these findings.

Despite the promising results, our study has limitations. First, the retrospective design with small sample size constraints further evaluation. Second, the absence of control groups fails to further validate safety and efficacy.

Conclusion

In conclusion, our study highlights the effectiveness of combining PPV with low-dose subretinal rt-PA injection and intravitreal anti-VEGF injection as a novel therapy for SMH. Significant clearance of hemorrhage from the macula with optimal visual outcomes after 6 months of treatment was observed. However, we noted instances of RPE degeneration, underscoring the importance of tailored treatment plans for individual patients. Furthermore, our findings emphasize the necessity of implementing appropriate combined therapies post-surgery to prevent SMH recurrence and preserve visual function effectively. Further prospective, intervention-based, multicenter, randomized controlled trials are warranted to validate the outcomes of this study and establish the efficacy of this combination therapy for SMH.